Indonesia's Retail and E-Commerce: A New Strategy Needed for Managing Marketing Channels

Indonesia’s digital economy continues to show rapid growth, with a growth rate reaching 14% and the e-commerce sector contributing 72% of the total Gross Merchandise Value (GMV). Amid increasingly fierce competition, artificial intelligence (AI) is now playing a more important role in interactions between brands and customers, from presenting personalised recommendations to supporting buying and selling interactions via chat on various messaging platforms.

However, behind the increasing adoption of AI, many retail players in Indonesia still face challenges in comprehensively understanding the customer journey. This is due to customer data that remains scattered and fragmented across various channels, making it difficult for companies to get a complete picture of consumer behaviour and needs.

For example, a customer might first discover a skincare brand through an Instagram advertisement, then contact the brand via Instagram direct message to find a product suitable for their skin needs, before finally completing a purchase on Shopee during a PayDay campaign using a discount voucher. Although the transaction is recorded in the marketplace, the prior interactions on Instagram are often not documented in the system. As a result, the brand may see the conversion as entirely driven by the marketplace, without realising the crucial role of social media interaction and customer consultation in influencing the purchase decision.

This challenge reflects a larger problem in Indonesian retail. Shopping activities via social media, transactions through WhatsApp, marketplaces, and physical stores still operate in separate systems that were not designed to be interconnected from the start. Each channel only records a part of the customer journey, from product discovery, conversation, and transaction, to repeat purchases, while the data generated often remains scattered across each platform.

Consequently, many retail companies and brand owners do not have a single, complete customer profile but rather several versions of customer data spread across different channels. This fragmentation limits AI’s ability to generate accurate insights, deliver more relevant interactions, and automate the customer journey effectively.

To bridge this gap, more than just adopting new technology is required. The industry needs a more integrated approach to data management and customer interaction coordination, so that every touchpoint—from social media, messaging apps, and marketplaces to physical stores—can be connected to form a more complete picture of the customer.

Firstly, unifying customer data from various channels is critical. Marketplaces like Shopee and Tokopedia only share a limited amount of buyer data with brands. As a result, many companies are essentially building an audience they cannot fully identify and customer relationships that are difficult to sustain in the long term. A customer who has made repeated transactions through a marketplace might still not be recorded in the brand’s CRM, loyalty programme, or AI-based personalisation system. The consequence is that many of the most active brands in e-commerce ironically have the most limited understanding of their best customers.

For instance, a customer who has made eight transactions through Shopee could remain unregistered in the brand’s CRM, loyalty programme, or personalisation system. When this same customer later contacts the brand via Instagram to submit a complaint or ask a question, the interaction often has to start from scratch, without the context of purchase history, without an understanding of customer preferences, and without continuity of the previously built relationship.

Unifying customer data means resolving the issue of customer identity before AI systems can be applied effectively. This requires infrastructure capable of collecting data from Instagram direct messages, WhatsApp Business, various marketplace dashboards, and point-of-sale systems in physical stores, and then integrating it into a single, complete customer profile even if the customer does not use the same identity across every channel. In this way, brands can build a single source of truth (SSOT) for customer data that is continuously updated at every interaction point and accessible to all teams that need it.

Secondly, brands need to shift towards a more predictive and contextual approach in customer interactions. Many brands only react after a customer takes action, for example by sending promotions after a transaction occurs or handling complaints only when the customer submits them. In reality, the biggest opportunities often arise before a purchase decision is made.

For example, a customer repeatedly views advertisements for running shoes on Instagram, then contacts the brand to ask about stock availability, but ultimately does not proceed with the purchase. Without connected data, these two activities appear as separate events. The marketing team only sees product interest, while the WhatsApp team sees an inquiry that did not lead to a transaction. Through data integration, these two signals can form a clearer picture of the customer’s purchase intent. AI can then recognise this pattern and trigger more relevant follow-ups, such as a more personal reminder, a special offer, or a suitable product recommendation at the time most likely to drive conversion.

A predictive approach also requires an understanding of local shopping behaviour patterns.